The Maximal Self-dissimilarity Interest Point Detector

We propose a novel interest point detector stemming from the intuition that image patches which are highly dissimilar over a relatively large extent of their surroundings hold the property of being repeatable and distinctive. This concept of contextual self-dissimilarity reverses the key paradigm of recent successful techniques such as the Local Self-Similarity descriptor and the Non-Local Means filter, which build upon the presence of similar rather than dissimilar patches. Moreover, our approach extends to contextual information the local self-dissimilarity notion embedded in established detectors of corner-like interest points, thereby achieving enhanced repeatability, distinctiveness and localization accuracy. As the key principle and machinery of our method are amenable to a variety of data kinds, including multi-channel images and organized 3D measurements, we delineate how to extend the basic formulation in order to deal with range and RGB-D images, such as those provided by consumer depth cameras.